DESCRIPTION Glaucoma is a group of diseases that results in a pathological loss of retinal ganglion cells (RGC) and irreversible vision loss. Although glaucoma is an optic neuropathy with characteristic optic nerve head (ONH) changes, risk of developing disease is not based on ONH structure, but factors including intraocular pressure (IOP), central corneal thickness, age, race and family history. In early disease, there is significant thinning of the optic nerve head (ONH) rim tissue that precedes RGC loss. We hypothesize that the early thinning of the ONH neuroretinal rim tissue (NRR) is related to changes in the glia and extracellular matrix, but not axonal content, which we will investigate using immunohistochemistry and 3D serial block-face scanning electron microscopy in the non- human primate experimental glaucoma model. We also hypothesize that the ONH NRR response to transient changes in IOP is a reflection of the NRR tissue composition, and predictive of the rate of RGC loss (SA1). Clinically, RGC content of the eye is assessed with non-invasive imaging using optical coherence tomography (OCT), for structure, and visual thresholds, for function. OCT structural measures have low variability and have revolutionized how glaucoma is assessed. However, the RGC correspondence to OCT measures is not known, and cannot be estimated from in vivo measures. In fact, the linear relationship for all OCT derived RGC measures is not correct. In SA2, the relationship between OCT derived measures of the circumpapillary retinal nerve fiber layer and ganglion cell inner plexiform thickness will be related to RGC content at all stages of neuropathy using rigorous histological methods. The goal of this aim is develop methods to estimate RGC content in the eye. For a disease that results in irreversible vision loss, it is important that visual function is also assessed accurately. In principal there should also be excellent correspondence between RGC content estimates from OCT measures and that from visual thresholds. Because structural measures are objective and less variable, it would be ideal to accurately predict vision using structural measures. However there is significant discrepancy between structural and functional measures. Some of the reasons for this disjunction is that visual function tests do not use appropriate spatial sampling and stimulus size. In these experiments we will investigate the relationship between RGC content and visual thresholds using higher spatial density and varying stimulus sizes (SA3). Our goal is to establish robust methods to predict visual function based on non-invasive structural imaging. Overall, these studies are designed to improve our understanding of disease pathophysiology and the ability to accurately monitor it in clinical practice.